Nowadays, Machine Learning and Deep Learning methods have become the state-of-the-art approach to solve data classification tasks. In order to use those methods, it is necessary to acquire and label a considerable amount of data; however, this is not straightforward in some fields, since data annotation is time consuming and might require expert knowledge. This challenge can be tackled by means of semi-supervised learning methods that take advantage of both labelled and unlabelled data. In this work, we present new semi-supervised learning methods based on techniques from Topological Data Analysis (TDA), a field that is gaining importance for analysing large amounts of data with high variety and dimensionality. In particular, we have created two semi-supervised learning methods following two different topological approaches.

Self-training (ST) is a straightforward and standard approach in semi-supervised learning, successfully applied to many machine learning problems. The performance of ST strongly depends on the supervised learning method used in the refinement step and the nature of the given data; hence, a general performance guarantee from a concise theory may become loose in a concrete setup. However, the theoretical methods that sharply predict how the performance of ST depends on various details for each learning scenario are limited. This study develops a novel theoretical framework for sharply characterizing the generalization abilities of the models trained by ST using the non-rigorous replica method of statistical physics. We consider the ST of the linear model that minimizes the ridge-regularized cross-entropy loss when the data are generated from a two-component Gaussian mixture. Consequently, we show that the generalization performance of ST in each iteration is sharply characterized by a small finite number of variables, which satisfy a set of deterministic self-consistent equations. By numerically solving these self-consistent equations, we find that ST's generalization performance approaches to the supervised learning method with a very simple regularization schedule when the label bias is small and a moderately large number of iterations are used.

After presenting SimCLR, a contrastive self-supervised learning framework, I decided to demonstrate another infamous method, called BYOL. Bootstrap Your Own Latent (BYOL), is a new algorithm for self-supervised learning of image representations. It does not explicitly use negative samples. Negative samples are images from the batch other than the positive pair. As a result, BYOL is claimed to require smaller batch sizes, which makes it an attractive choice.

Following the success of supervised learning, semi-supervised learning (SSL) is now becoming increasingly popular. SSL is a family of methods, which in addition to a labeled training set, also use a sizable collection of unlabeled data for fitting a model. Most of the recent successful SSL methods are based on pseudo-labeling approaches: letting confident model predictions act as training labels. While these methods have shown impressive results on many benchmark datasets, a drawback of this approach is that not all unlabeled data are used during training. We propose a new SSL algorithm, DoubleMatch, which combines the pseudo-labeling technique with a self-supervised loss, enabling the model to utilize all unlabeled data in the training process. We show that this method achieves state-of-the-art accuracies on multiple benchmark datasets while also reducing training times compared to existing SSL methods. Code is available at https://github.com/walline/doublematch.

Crime has become a major concern in many cities, which calls for the rising demand for timely predicting citywide crime occurrence. Accurate crime prediction results are vital for the beforehand decision-making of government to alleviate the increasing concern about the public safety. While many efforts have been devoted to proposing various spatial-temporal forecasting techniques to explore dependence across locations and time periods, most of them follow a supervised learning manner, which limits their spatial-temporal representation ability on sparse crime data. Inspired by the recent success in self-supervised learning, this work proposes a Spatial-Temporal Hypergraph Self-Supervised Learning framework (ST-HSL) to tackle the label scarcity issue in crime prediction. Specifically, we propose the cross-region hypergraph structure learning to encode region-wise crime dependency under the entire urban space. Furthermore, we design the dual-stage self-supervised learning paradigm, to not only jointly capture local- and global-level spatial-temporal crime patterns, but also supplement the sparse crime representation by augmenting region self-discrimination. We perform extensive experiments on two real-life crime datasets. Evaluation results show that our ST-HSL significantly outperforms state-of-the-art baselines. Further analysis provides insights into the superiority of our ST-HSL method in the representation of spatial-temporal crime patterns. The implementation code is available at https://github.com/LZH-YS1998/STHSL.

One of the first steps in many text-based social science studies is to retrieve documents that are relevant for the analysis from large corpora of otherwise irrelevant documents. The conventional approach in social science to address this retrieval task is to apply a set of keywords and to consider those documents to be relevant that contain at least one of the keywords. But the application of incomplete keyword lists risks drawing biased inferences. More complex and costly methods such as query expansion techniques, topic model-based classification rules, and active as well as passive supervised learning could have the potential to more accurately separate relevant from irrelevant documents and thereby reduce the potential size of bias. Yet, whether applying these more expensive approaches increases retrieval performance compared to keyword lists at all, and if so, by how much, is unclear as a comparison of these approaches is lacking. This study closes this gap by comparing these methods across three retrieval tasks associated with a data set of German tweets (Linder, 2017), the Social Bias Inference Corpus (SBIC) (Sap et al., 2020), and the Reuters-21578 corpus (Lewis, 1997). Results show that query expansion techniques and topic model-based classification rules in most studied settings tend to decrease rather than increase retrieval performance. Active supervised learning, however, if applied on a not too small set of labeled training instances (e.g.

AI has classically come in three forms, supervised learning, unsupervised learning, and reinforcement learning. Supervised learning is where AI is given many example scenarios and the right answer for each one (such as images labeled as Cat or Dog). Unsupervised learning has been traditionally where AI learns to group items together by similarity (clustering), without explicit labels. Reinforcement learning is where AIs try out strategies (such as in a game) and attempt to optimize a reward function (such as points in the game). Many commercial AIs are based on supervised learning.

This story will explore how we can reason from and model graphs using labels via Supervised and Semi-Supervised Learning. I'm going to be using a MET Art Collections dataset that will build on my previous parts on Metrics, Unsupervised Learning, and more. Be sure to check out the previous story before this one to keep up on some of the pieces as I won't cover all concepts again in this one: The easiest approach to conduct Supervised Learning is to use graph measures as features in a new dataset or in addition to an existing dataset. I have seen this method yield positive results for modeling tasks, but it can be really dependent on 1. how you model as a graph (what are the inputs, outputs, edges, etc.) and 2. which metrics to use. Depending on the prediction task, we could compute node-level, edge-level, and graph-level metrics.

Self-supervised learning (SSL) is gaining a larger foothold in the world of machine learning (ML). As learning models are refined and expanded, machines that teach themselves, understand context and are able to fill in the blanks where there are holes in the information are the next step. Machines are taught to analyze, predict and advise on possible outcomes. Supervised learning - Practitioners train the machine on inputs paired with labelled outputs, teaching it to make associations. Example: A shape with three sides is labelled triangle .

Yann LeCun is the Chief AI Scientist at Meta, professor at NYU, Turing Award winner, and one of the seminal researchers in the history of machine learning. Please support this podcast by checking out our sponsors: – Public Goods: https://publicgoods.com/lex and use code LEX to get $15 off – Indeed: https://indeed.com/lex Books and resources mentioned: Self-supervised learning (article): https://bit.ly/3Aau1DQ SUPPORT & CONNECT: – Check out the sponsors above, it's the best way to support this podcast – Support on Patreon: https://www.patreon.com/lexfridman On some podcast players you should be able to click the timestamp to jump to that time.